Research Description

My research challenges one of the central assumptions of contemporary organizations: that bringing different people into the room necessarily brings different ideas into play. Organizations today invest enormous resources in diversity, connectivity, and expertise because these are presumed to fuel innovation. My work shows that they can just as easily do the opposite. I study how the very structures designed to expand participation can produce intellectual closure, shaping not only who gets included, but what kinds of knowledge inclusion is allowed to generate. My first major project develops a theory of the diversity paradox: demographically diverse teams can become cognitively homogeneous. I argue that visible difference creates uncertainty, and that actors often resolve this uncertainty by seeking familiarity along less visible dimensions, such as training, expertise, professional networks, and problem frames. I call this process compensatory cognitive homophily. Using data on more than four million patent teams, I employ natural language processing to locate inventors in a knowledge space and estimate the cognitive distance among collaborators. The findings reveal a hidden liability of diversity efforts: representation can expand while the range of knowledge brought into collaboration narrows. My second major project extends this agenda to medicine, using claims data from billions of healthcare encounters to examine how physician networks govern the diffusion of expertise, the adoption of new practices, and the evolution of knowledge. Together, these projects advance a theory of innovation as a social and cognitive sorting process. Networks are not neutral channels through which ideas flow; they are gatekeeping systems that decide which expertise matters, which knowledge travels, and which combinations of ideas become thinkable. My research therefore pushes organizational and management theory beyond asking whether organizations are diverse or connected, and toward a more fundamental question: when do organizations actually allow differences to make a difference?